With the rapid growth of Indian Film music industry huge volume of Hindi songs are produced every day. Numerous of lyricists, singers and artists are involved in production of songs from different genres. With evolved research in field of Music Information Retrieval analysis on Hindi songs emerged nowadays, but open source dataset availability is less. Thus, proposed technique to pre-pare dataset by crawling lyrics based on song title from the webpage will be extracted by using classical naïve based string matching algorithm. Metadata extraction procedure will be useful to extract various Meta information of song paired with music, which will be useful to prepared repository with essential song metadata. The prepared dataset given Hindi transliterate name “Sanchay” means a repository will helpful in area of mu-sic segmentation, digital signal processing, song feature extraction, emotion classification, lyrics mining, playlist generation, artist-genre paired clustering with commercial applications too.

music information retrieval, lyrics mining, signal processing, metadata, genre, text transliterate